The control of cell number in multicellular eukaryotes represents a balance between cell proliferation and cell death. Although a great deal has been learned in recent years about the regulation of cell proliferation, relatively little is known about the regulation of cell death (Ellis et al., 1991; Raff, 1992). Recently, attention has begun to focus on the mechanisms that regulate programmed cell death (apoptosis) (Williams, 1991). Apoptosis is an active process by which many cells die during development and self-maintenance in complex eukaryotes (Kerr et al., 1972). Cell death by apoptosis occurs when a cell activates an internally encoded suicide program as a result of either extrinsic or intrinsic signals. Apoptotic cell death is characterized by plasma membrane blebbing, cell volume loss, nuclear condensation, and endonucleolytic degradation of DNA at nucleosomal intervals (Wyllie et al., 1980).
Two of the best studied vertebrate systems in which programmed cell death plays a role are neural and lymphoid development. During T cell development in the thymus, each individual T cell precursor generates a unique T cell antigen receptor (TCR) by combinatorial rearrangement of TCR gene segments and the cell subsequently undergoes a series of selection processes (Blackman et al., 1990; Rothenberg, 1992). T cells expressing autoreactive TCRs are deleted by apoptosis as a result of negative selection (Murphy et al., 1990). Other cells undergo positive selection through interaction with self-encoded major histocompatibility complex (MHC) molecules expressed on thymic stromal cells, a process which prevents programmed cell death and results in the subsequent MHC-restriction of the mature T cell repertoire. An additional set of thymic cells die as a result of neglect, the absence of either negative or positive selection. Extensive cell death also occurs in the developing nervous system (Cowan et al., 1984; Davies, 1987; Oppenheim, 1991). Following an initial expansion of neurons during development, a significant reshaping of neural structures occurs as a result of the establishment of synaptic interactions. During this reshaping period, the survival of neurons is determined by their supply of neurotrophic growth factors. Cells that become growth-factor deficient die by apoptosis. Once synaptic connections are established, the surviving neurons develop into post-mitotic cells with extended life spans. Thus, programmed cell death plays an essential role in lymphoid development by removing autoreactive T cells and within the nervous system by facilitating the establishment of effective synaptic networks.
Because of the importance of programmed cell death to these developmental processes, considerable interest has arisen in genes that are capable of regulating apoptosis. One of the most important advances in the understanding of the regulation of apoptotic cell death in vertebrates has come from studies of the oncogene bcl-2. bcl-2 was originally cloned from the breakpoint of a t(14; 18) translocation present in many human B cell lymphomas (Cleary et al., 1986; Tsujimoto et al., 1986). This translocation results in the deregulated expression of the bcl-2 gene as result of its juxtaposition with the immunoglobulin heavy chain gene locus (Bakhshi et al., 1985). In vitro, BCL-2 (the gene product of bcl-2) has been shown to prevent apoptotic cell death in cultured cells which are deprived of growth factors (Vaux et al., 1988; Hockenbery et al., 1990; Nuñez et al., 1990; Borzillo et al., 1992; Garcia et al., 1992). However, BCL-2 is not able to block apoptosis in all cells induced by cytokine deprivation or receptor-mediated signalling. For example, BCL-2 prevents apoptosis in hematopoietic cell lines dependent on certain interleukins (IL) IL-3, IL-4, or GM-CSF but it fails to prevent other cell lines from apoptosis following IL-2 or IL-6 deprivation (Nuñez et al., 1990). Overexpression of BCL-2 also fails to prevent antigen receptor-induced apoptosis in some B cell lines (Cuende et al., 1993). In vivo, BCL-2 prevents many, but not all, forms of apoptotic cell death that occur during lymphoid (Sentman et al., 1991; Strasser et al., 1991a; Strasser et al., 1991b; Seigel et al., 1992) and neural (Allsop et al., 1993) development. Expression of a bcl-2 transgene can prevent radiation- and calcium ionophore-induced apoptotic cell death in thymocytes, but does not inhibit the process of negative selection (Sentman et al., 1991; Strasser et al., 1991a). Similarly, overexpression of bcl-2 can prevent apoptosis in neurons dependent on nerve growth factor, but not neurons dependent upon ciliary neurotrophic factor. (Allsop et al., 1993) These results suggest the existence of multiple independent intracellular mechanisms of apoptosis, some of which can be prevented by BCL-2 and others which are unaffected by this gene. Alternatively, these additional pathways may involve proteins that differentially regulate BCL-2 function.